Hacker News: lovasoa

New comment by lovasoa in "Caltech Mathathon – first hackathon ever devoted to research level mathematics"

lovasoa — Mon, 07 Sep 2026 20:42:51 +0000

steam engines solved a problem people had

New comment by lovasoa in "Any Human Ever – One life, drawn at random from all who have ever lived"

lovasoa — Thu, 03 Sep 2026 17:56:11 +0000

In 429 CE, Devadatta was born on the Deccan Plateau, near present-day Hyderabad. His hamlet was home to about 100 people. Among six children, he was the second. Devadatta died of neonatal tetanus or sepsis within days of birth.

Hardcore

New comment by lovasoa in "Maximizing the value of your Claude Code sessions"

lovasoa — Sat, 15 Aug 2026 06:47:44 +0000

I don't understand why changing effort levels busts the cache. Couldn't effort levels be a decoding-only thing where they just change the probability of the token? Are they literally adding a hidden system prompt that says "effort level: $level" ?

New comment by lovasoa in "Qwen3.8-Max: A New Bar for Coding and Cowork"

lovasoa — Mon, 03 Aug 2026 06:58:11 +0000

the writing has been on the wall for fifteen years, and yet the number of professional drivers has only increased since then...

New comment by lovasoa in "Gemini Robotics 2 brings whole body intelligence to robots"

lovasoa — Thu, 30 Jul 2026 17:20:34 +0000

Of course we'll rent them! I don't even want to keep a robot in my flat all the time.

But that means speed does matter after all. The renting company will need to clean as many flats as possible with a single robot in a single day.

New comment by lovasoa in "Anatomy of a Frontier Lab Agent Intrusion: A Timeline of the July 2026 Incident"

lovasoa — Thu, 30 Jul 2026 07:15:18 +0000

I was skeptical after last week's announcements, and I have to say I'm impressed now. Over the last two days, I reproduced the entire chain of components involved, and the exploits at play, and even though none of the exploits are crazy smart, the long series of pivots demonstrates a level of agency I didn't think today's LLMs had.

https://github.com/lovasoa/hf-ctf

OpenAI's HuggingFace Hack Reconstitution

lovasoa — Wed, 29 Jul 2026 21:47:49 +0000

Article URL: https://github.com/lovasoa/hf-ctf

Comments URL: https://news.ycombinator.com/item?id=49103506

Points: 2

# Comments: 0

New comment by lovasoa in "Fast Remediation Is the New Trust Model (JFrog and OpenAI Zero-Day Findings)"

lovasoa — Tue, 28 Jul 2026 17:00:14 +0000

The timeline is indeed a little bit fuzzy, I haven't found a precise chronology, neither from HuggingFace nor from OpenAI. HF says the hack happened "over a weekend", so probably July 11th-12th. Do you think it took OpenAI a week to realize what happened ?

Maybe when HF published their blog post on the 15th, OpenAI already knew something had happened, had started to investigate, and already reported the issue to JFrog ? But looking at your other comment in the thread, I agree that CVE-2026-65925 and CVE-2026-66014 are better candidates.

Taking a step back, so many basic vulnerabilities in a security-oriented product just makes the headline "agent autonomously escaped containment" sound a little less spectacular.

New comment by lovasoa in "Fast Remediation Is the New Trust Model (JFrog and OpenAI Zero-Day Findings)"

lovasoa — Tue, 28 Jul 2026 13:57:28 +0000

There is one thing that is still unclear to me after reading a lot about the hack:

* were the ExploitGym solutions actually available somewhere inside huggingface's private datasets ?

* was the model really trying to extract the solutions ? or had some sub-agent drifted enough from the original context that it was not even trying to solve the initial challenge ? that would look much worse for OpenAI, PR-wise.

New comment by lovasoa in "Fast Remediation Is the New Trust Model (JFrog and OpenAI Zero-Day Findings)"

lovasoa — Tue, 28 Jul 2026 13:49:20 +0000

What they conveniently omit in the blog post is what the vulnerability was: it seems like they renewed JWTs without checking the signature at all ! You could write arbitrary info in an old token, and get it signed without any verification.

https://www.youtube.com/watch?v=q2KCrmQz9WE

New comment by lovasoa in "I rewrote PostHog's SQL parser, 70x faster, while barely looking at the code"

lovasoa — Wed, 24 Jun 2026 19:57:00 +0000

The thing I would have liked to know is why they don't use an existing fast SQL parser. Was being slightly incompatible with all existing SQL dialects a product requirement?

New comment by lovasoa in "OpenAI unveils its first custom chip, built by Broadcom"

lovasoa — Wed, 24 Jun 2026 18:19:19 +0000

3) The engineers working on the chip used ChatGPT from time to time.

New comment by lovasoa in "Silurus/ooxml: Pixel-faithful Office documents, rendered in the browser"

lovasoa — Sun, 07 Jun 2026 19:41:04 +0000

I tried a few PPTX files from consulting firms that are available online. The rendering does not seem to be truly pixel-perfect, but all of them were quite readable and had a good layout, which is already an impressive feat.

New comment by lovasoa in "Shipping a laptop to a refugee camp in Uganda"

lovasoa — Sat, 23 May 2026 21:37:06 +0000

The details are interesting: https://en.wikipedia.org/wiki/Eskom#Corruption

New comment by lovasoa in "Redis array: short story of a long development process"

lovasoa — Mon, 04 May 2026 16:42:51 +0000

How much faster/slower are you with that process compared to writing code yourself?

New comment by lovasoa in "Is anybody else bored of talking about AI?"

lovasoa — Tue, 24 Mar 2026 21:35:20 +0000

I use the latest codex with gpt5.4 and Claude opus every day. they hallucinate every day. If you think they don't, you are probably being gaslighted by the models.

New comment by lovasoa in "“Car Wash” test with 53 models"

lovasoa — Tue, 24 Feb 2026 13:58:25 +0000

I was interested in the human results, so I had an llm build a visualization for them: https://codepen.io/lovasoaaa/pen/QwKWGBd

You can see that 17% of answers come from India alone and that software developers got below average results, for instance.

New comment by lovasoa in "Flux 2 Klein pure C inference"

lovasoa — Sun, 18 Jan 2026 22:25:25 +0000

The author of this project is also the author of redis. He knows what he is doing.

Running inference for a model, even when you have all the weights, is not trivial.

New comment by lovasoa in "The struggle of resizing windows on macOS Tahoe"

lovasoa — Sun, 11 Jan 2026 22:46:19 +0000

I use linux at home (with a HiDPI screen) and MacOS for work. The screen works well with both computers. I mostly just use a text editor, a browser, and a terminal though.

Linux has bugs, bug MacOS does too. I feel like for a dev like me, the linux setup is more comfortable.

SQLPage 0.38: transform SQL queries into web UIs for any DB

lovasoa — Tue, 14 Oct 2025 22:17:41 +0000

Article URL: https://github.com/sqlpage/SQLPage/releases/tag/v0.38.0

Comments URL: https://news.ycombinator.com/item?id=45585717

Points: 2

# Comments: 0